Abstract
Bauxsol is a chemico-physically modified product of red mud and is a promising material for the removal and recovery of phosphorus from wastewater. In this study, response surface methodology (RSM) and artificial neural network (ANN) were employed to develop prediction models and also to investigate the interactions of independent experimental factors for phosphorus adsorption onto acid-activated Bauxsol. The experimental results indicated that HCl activation was effective to improve the adsorption capacity of Bauxsol. The maximum adsorption capacity of acid-activated Bauxsol was 55.72 mg/g (as P) with HCl concentration of 10.20 mol/L, temperature of 41.00 degrees C, and time of 5.60 h, which increased by 10.53 and 6.62 times compared with the raw red mud and Bauxsol before acid activation, respectively. The relative importance of HCl concentration in RSM and ANN models was 51.78 and 54.25 %, respectively, which illustrated that HCl concentration played the predominant role on improving the adsorption capacity of Bauxsol. The predictive capability of RSM and ANN models was compared, and the results showed that both models provided excellent predictions with R-2>0.93. However, the ANN model showed the superiority over RSM for estimation capability.